add : manual sweep

image_ref/main_sweep.py

-import time
-
-import wandb as wdb
+import random
+import numpy as np
+from sweep_train import run_duo
 
 if __name__ == '__main__':
     sweep_configuration = {
@@ -10,7 +10,7 @@ if __name__ == '__main__':
         "parameters": {
             "epoches":{"value": 50},
             "eval_inter":{"value": 1},
-            "noise_threshold": {"distribution" : "log_uniform_values", "max": 10000., "min": 0.0001},
+            "noise_threshold": {"distribution" : "log_uniform_values", "max": 10000., "min": 1},
             "lr": {"distribution" : "log_uniform_values", "max": 0.01, "min": 0.0001},
             "batch_size": {"value": 64},
             "positive_prop": {"distribution" : "uniform","max": 95., "min": 5.},
@@ -21,15 +21,21 @@ if __name__ == '__main__':
             "dataset_val_dir": {"value": "data/processed_data_wiff/npy_image/test_data"},
             "dataset_ref_dir": {"values": ["image_ref/img_ref","image_ref/img_ref_count_th_10","image_ref/img_ref_count_th_5"]},
         },
-        "controller":{
-            "type": "local"},
+        "max_iter": 10
     }
 
-    sweep_id = wdb.sweep(sweep_configuration)
+    for i in range(sweep_configuration["max_iter"]):
+        run_config={}
+        for p,v in sweep_configuration["parameters"].items() :
 
-    # Start the local controller
-    sweep = wdb.controller(sweep_id)
-    while not sweep.done():
-        sweep.print_status()
-        sweep.step()
-        time.sleep(5)
+            if "value" in v:
+                run_config[p]=v["value"]
+            elif "values" in v:
+                run_config[p] = random.choice(v["values"])
+            elif "distribution" in v:
+                if v["distribution"]=="uniform":
+                    run_config[p] = random.uniform(v["min"],v["max"])
+                elif v["distribution"]=="log_uniform_values":
+                    run_config[p] = np.exp(random.uniform(np.log(v["min"]), np.log(v["max"])))
+        print('Launching run')
+        run_duo(run_config)

image_ref/sweep_train.py

@@ -6,7 +6,7 @@ import torch.nn as nn
 from model import Classification_model_duo_contrastive
 import torch.optim as optim
 
-def train_duo(model, data_train, optimizer, loss_function, epoch, wandb):
+def train_duo(model, data_train, optimizer, loss_function, epoch):
     model.train()
     losses = 0.
     acc = 0.
@@ -40,7 +40,7 @@ def train_duo(model, data_train, optimizer, loss_function, epoch, wandb):
     return losses, acc
 
 
-def val_duo(model, data_test, loss_function, epoch, wandb):
+def val_duo(model, data_test, loss_function, epoch):
     model.eval()
     losses = 0.
     acc = 0.
@@ -94,15 +94,15 @@ def run_duo(args):
 
     print('Wandb initialised')
     # load data
-    data_train, data_val_batch, data_test_batch = load_data_duo(base_dir_train=args.dataset_train_dir,
-                                                                base_dir_val=args.dataset_val_dir,
+    data_train, data_val_batch, data_test_batch = load_data_duo(base_dir_train=args['dataset_train_dir'],
+                                                                base_dir_val=args['dataset_val_dir'],
                                                                 base_dir_test=None,
-                                                                batch_size=args.batch_size,
-                                                                ref_dir=args.dataset_ref_dir,
-                                                                positive_prop=args.positive_prop, sampler=args.sampler)
+                                                                batch_size=args['batch_size'],
+                                                                ref_dir=args['dataset_ref_dir'],
+                                                                positive_prop=args['positive_prop'], sampler=args['sampler'])
 
     # load model
-    model = Classification_model_duo_contrastive(model=args.model, n_class=2)
+    model = Classification_model_duo_contrastive(model=args['model'], n_class=2)
     model.float()
     # move parameters to GPU
     if torch.cuda.is_available():
@@ -118,15 +118,15 @@ def run_duo(args):
     val_loss = []
     # init training
     loss_function = nn.CrossEntropyLoss()
-    if args.opti == 'adam':
-        optimizer = optim.Adam(model.parameters(), lr=args.lr)
+    if args['opti'] == 'adam':
+        optimizer = optim.Adam(model.parameters(), lr=args['lr'])
     # train model
-    for e in range(args.epoches):
-        loss, acc = train_duo(model, data_train, optimizer, loss_function, e, args.wandb)
+    for e in range(args['epoches']):
+        loss, acc = train_duo(model, data_train, optimizer, loss_function, e)
         train_loss.append(loss)
         train_acc.append(acc)
-        if e % args.eval_inter == 0:
-            loss, acc, acc_contrastive = val_duo(model, data_val_batch, loss_function, e, args.wandb)
+        if e % args['eval_inter'] == 0:
+            loss, acc, acc_contrastive = val_duo(model, data_val_batch, loss_function, e)
             val_loss.append(loss)
             val_acc.append(acc)
             val_cont_acc.append(acc_contrastive)
@@ -134,6 +134,4 @@ def run_duo(args):
 
 
 if __name__ == '__main__':
-    config = wdb.config
-    print(config)
-    run_duo(config)
\ No newline at end of file
+    pass
\ No newline at end of file